Signals without context
Historians and dashboards expose measurements, but rarely connect them to design intent, process relationships or physical cause.
Sentinel OS · Industrial operating intelligence
See the state.
Understand why.
Act with confidence.
Sentinel OS fuses live plant data with first-principles physics to build a continuously updated model of industrial operations—turning fragmented signals into trusted understanding, decision support and progressively autonomous action.
Feed
System
System
Process
Unit A
Unit A
Pump
P-204
P-204
Utility
Loop
Loop
Separator
S-17
Physics-constrained state graph · 1,248 live signals
S-17
State update Pump P-204 cavitation margin decreasing
Physics model Mass balance reconciled across Unit A
Constraint Separator S-17 approaching hydraulic limit
Decision Alternative operating plan evaluated
Autonomy Approved optimisation envelope active
Edge sync Site model healthy · latency 11ms
State update Pump P-204 cavitation margin decreasing
Physics model Mass balance reconciled across Unit A
Constraint Separator S-17 approaching hydraulic limit
Decision Alternative operating plan evaluated
Autonomy Approved optimisation envelope active
Edge sync Site model healthy · latency 11ms
01 · ObserveConnect every relevant signal
02 · UnderstandInfer the true physical state
03 · DecideEvaluate actions and constraints
04 · ActCoordinate people and systems
The operating problem
Industrial sites have thousands of signals.
They still lack a shared model of reality.
They still lack a shared model of reality.
Decisions without a common state
Operations, reliability and engineering teams often interpret the same process through different tools and assumptions.
Automation without understanding
Rigid control logic can execute actions, but it cannot reliably reason about changing plant conditions beyond predefined rules.
A physics-native operating system
A common operational picture.
Grounded in physics.
Sentinel OS continuously reconciles what sensors observe, what engineering models predict and what operational constraints permit. The result is a living state model that teams and autonomous systems can use to understand the plant and determine what should happen next.
Connect
SCADA, historians, PLCs, laboratory data, CMMS, edge devices and engineering records.
Contextualise
Map signals to assets, process streams, operating modes, design limits and dependencies.
Estimate state
Align time, reconcile data quality and infer physical variables that cannot be measured directly.
Apply physics
Run mass and energy balances, fluid, thermal, mechanical and process-specific models.
Decide
Diagnose causes, test scenarios and rank actions by expected impact, risk and confidence.
Coordinate action
Guide operators, integrate workflows and enable bounded machine execution over time.
One source of operational truth
Measured. Estimated.
Expected. Explained.
Sentinel separates what the plant reports from what is actually happening.
Every insight retains its evidence: observed signals, inferred state, physics expectation, deviation, likely causes, uncertainty and recommended action. Engineers can inspect the reasoning before deciding whether to act.
This is the foundation for trust: not a black-box score, but an auditable operational state.
ObservedDischarge pressure
Current value6.42 bar
Estimated stateCavitation margin
StatusNarrowing
Expected stateAt current duty
Deviation−11.8%
Likely causeInlet condition change
Confidence86%
Recommended actionReduce speed 3%
Constraint checkApproved
Core capabilities
Built for decisions,
not another dashboard.
Sentinel OS turns a shared state model into operational capabilities that can start with one asset or process and expand across the site.
Real-time process state
Build a unified view of assets, process streams, constraints and operating modes from fragmented plant systems.
- Asset and process graph
- State estimation
- Data quality and uncertainty
Physics-based diagnosis
Explain deviations through physical mechanisms instead of relying only on thresholds or historical correlation.
- Root-cause hypotheses
- Operating-envelope analysis
- Cross-asset effects
Scenario and action engine
Evaluate available actions before they are applied and rank them by impact, confidence and operational risk.
- What-if simulation
- Constraint checking
- Expected-value ranking
Human-machine workflows
Deliver recommendations into existing operational workflows, with approvals, auditability and feedback.
- Role-based decision queues
- Operator acknowledgement
- Action outcome capture
Resilient site runtime
Run critical models locally, maintain state through connectivity loss and synchronise when network access returns.
- Low-latency inference
- Disconnected operation
- Cloud-edge synchronisation
Bounded optimisation
Allow approved actions to execute within explicit physical, safety and business constraints as trust grows.
- Approved action envelopes
- Human override
- Full decision trace
A responsible path to autonomy
Autonomy is earned.
Not switched on.
Sentinel OS is designed to sit above existing control and safety systems. Teams can progress from visibility to bounded execution only as models are validated and operating confidence grows.
Observe
Unify live signals and engineering context into a common operating picture.
Explain
Identify physical causes, deviations and developing process conditions.
Recommend
Propose actions with evidence, confidence and expected operational effect.
Approve & execute
Trigger actions after human approval through existing control and workflow systems.
Autonomously optimise
Continuously act inside validated constraints, with oversight, traceability and override.
Brownfield-first deployment
Start with one operational decision.
Scale to an operating layer.
Sentinel OS complements existing SCADA, historian, DCS and safety systems. It can begin as a read-only intelligence layer and expand without a rip-and-replace programme.
Choose the decision
Identify one costly, repeatable decision with available data and a measurable operational outcome.
Build the state model
Map data, assets, process relationships, design intent and approved operating constraints.
Run alongside operations
Compare Sentinel’s understanding and recommendations with engineering judgement and actual outcomes.
Reuse the foundation
Add models, assets, process units and sites while preserving a shared architecture and decision history.
Design partner programme
Build the first Sentinel deployment with us.
We are working with a small number of industrial organisations to validate Sentinel OS against real operational decisions. The strongest starting point is a process where teams already have data, but still rely heavily on engineering judgement to understand what is happening and determine what to do next.
One site or process unit
One high-value operational decision
Existing plant data first
Physics model validation
Measured operational outcome
Progressive path to action